Sie befinden sich hier:
  1. Start
  2. Challenge
  3. Net Working Capital Optimization
Evonik Industries AG

Net Working Capital Optimization

Detailed description of challenge

  • The optimization of NWC (Net Working Capital), which is comprised of Inventory, Accounts Receivable and Accounts Payable, is difficult to perform manually as there are large amounts of data with multiple influencing factors and levers that determine their level. Factors that contribute to the complexity include, inter alia: multiple production sites, legal entities, countries, currencies, high number of SKUs (Stock Keeping Units), minimum production quantities, production sequencing where certain processes must take place before others.
  • The objective of the challenge is to leverage an (existing) application that can visualize, forecast, and simulate (when adjusting levers) the NWC positions by using existing process data in SAP (and sometimes other tools), and then using ML (Machine Learning) make proposals for optimization actions. The application requires a hybrid approach (domain know-how/process rules, ML).

First initiatives within company

  • Activities already undertaken include:

    • – Investigation into off the shelf tools. Until now, no complete application that consolidates all three elements of NWC was identified. Simulation is only possible where all three elements are present in one interface and this was not possible.
    • – Individual analytics per process are performed manually, with limited or no ML capabilities.

Need for support

  • The expectation for the solution is to leverage an (existing) application that can:

    • – Access the end-to-end data of the three elements of NWC
    • – Provide an interface for viewing the data, and drilling down into the data
    • – Provide an interface for viewing and modifying individual parameters
    • – By leveraging ML and AI functionalities:
      • – Identify and propose optimization possibilities
      • – Learn from past decisions and contributed expert knowledge (e.g. defer for a defined selected time period, action/execute a decision for a defined time period, do not execute/action at all)
      • – Learn from business basics (e.g. negative inventory levels are not possible)
    • – Feedback Loop: Checking the success of the implementation of optimization proposals

Diese Challenge ist CLOSED!

— Bewerbungen derzeit nicht möglich.

Tags

Machine LearningNet Working Capital
Juni 21, 2022

 

Evonik Industries AG

Evonik is one of the world leaders in specialty chemicals. The company is active in more than 100 countries around the world and generated sales of €13.1 billion and an operating profit (adjusted EBITDA) of €2.15 billion in 2019. Evonik goes far beyond chemistry to create innovative, profitable and sustainable solutions for customers. More than 32,000 employees work together for a common purpose: We want to improve life, today and tomorrow.

https://corporate.evonik.de/de